LOUSE FEEDING EXPERIMENTS 49 



of any case, during the time it was under our observation, gives 

 clinical ground for suspecting the existence of trench fever. 

 None of the cases were related nor came from a common ad- 

 dress. It is of interest to note that in Box LII, Experiment 40, 

 though fed only upon patients used in feeding the doubly in- 

 fected lice, neither Rickettsia prowazeki nor Rickettsia pediculi 

 appeared. 



The occasional occurrence of Rickettsia pediculi was antici- 

 pated by us because of our previous experience with the lice 

 from the Warsaw bath-house from which Mr. Bacot pre- 

 sumably became infected with trench fever. 



Micro-organisms other than rickettsia did not appear in any 

 of the fifty-two experiments as the result of the feeding upon 

 typhus patients. The cocco-bacillus, found in the genital tract 

 of one louse from Box vi, Experiment 5, was familiar to us from 

 our initial controls of the English stock lice. The lice of Box 

 vi, however, were of the American stock and this was the first 

 instance of the occurrence of this bacterium in the American 

 stock. 



We have found it to be more difficult to infect lice with 

 Rickettsia prowazeki than is apparently indicated in the ac- 

 counts of da Rocha-Lima (1916, p. 29). Even in heavily in- 

 fected boxes in the later half of our work there was always a 

 varying percentage of lice in which rickettsia could not be 

 demonstrated, and in Box LII not one of the twenty-one lice 

 recovered and examined, twelve by serial sections, nine by 

 smears, showed rickettsia. It is certain that these negative lice 

 were exposed equally with their companions to rickettsia in- 

 fection. No suggestion can be advanced by us in explanation 

 for the absence of rickettsia in these lice; but, as shown below, 

 we have proved that only a similar proportion of the lice fed 

 upon typhus patients acquire the virus of typhus. In Box LII a 

 few lice showed appearances in cells of the alimentary tract 

 which later may be proved to be indications of early infection 

 with Rickettsia prowazeki. 



In these experiments no attempt was made to ascertain the 

 infectivity of patients at various stages of the disease as shown 



